B. fragilis toxin production

Some strains of Bacteroides fragilis carry the genetic instructions for making a toxin that can damage the colon lining. Experiments have established several ways this toxin causes harm, and human studies have linked these strains to diarrhea and colorectal cancer. But carrying the genetic instructions doesn't establish that the toxin is causing damage in your gut.

This stool result estimates toxin-producing potential from microbial DNA. Its use for predicting disease or guiding preventive treatment remains exploratory. A result cannot establish whether you have polyps, cancer or an inflammatory bowel disease flare.

What This Result Represents

The test uses shotgun metagenomic sequencing. This method reads DNA from the organisms in your stool and identifies genes associated with particular functions. Here, the function is the capacity to produce Bacteroides fragilis toxin.

The toxin protein is called BFT or fragilysin, and the bacterial gene that encodes it is called bft. Strains carrying this gene are called enterotoxigenic B. fragilis, often shortened to ETBF. These names refer to a protein, a gene and a group of bacterial strains, respectively.

The reported percentage is a relative estimate of toxin-producing potential within the sampled microbial community. It doesn't measure toxin protein concentration, the rate of toxin release or damage to your colon. A separate result for total B. fragilis abundance includes strains with and without the toxin gene.

How the Toxin Can Damage the Colon Lining

BFT is a protein-cutting enzyme that needs zinc to function. In experiments with intestinal cells, exposure causes cleavage of E-cadherin. E-cadherin helps neighboring cells stay attached, so its disruption can weaken the intestinal barrier.

Cell studies also show changes in beta-catenin signaling, which helps control cell growth, and activation of inflammatory responses. These include increased release of interleukin-8. This signaling protein attracts immune cells. The mechanisms are supported by laboratory experiments; finding toxin-producing potential in stool doesn't demonstrate that these processes are occurring in your colon.

Other experiments connect toxin exposure to increased activity of spermine oxidase. This enzyme can generate reactive molecules that damage DNA. Studies in mice susceptible to intestinal tumors show that toxin-producing strains can promote inflammation and tumor formation. Those findings support a possible role in human cancer, but they don't establish how much risk a particular stool result carries.

Colon Polyps and Colorectal Cancer

Human studies often find toxin-producing strains more frequently in people with colorectal cancer. A 2025 review included 14 studies, with 13 contributing to the comparison between cancer and healthy control groups. The pooled odds of detecting ETBF were about two and a half times higher in the cancer groups.

That analysis combined different detection methods and both stool and colon tissue samples. It describes an association with existing cancer, not the probability that someone with this stool result will develop cancer later. Most studies used targeted gene testing rather than this metagenomic percentage.

In a French study of 129 people, targeted stool testing detected ETBF in 56.5% of those with adenomas and 24% of those with a normal colonoscopy. ETBF was the only one of the five bacterial markers studied that was significantly more common in the adenoma group. The other bacteria were also detectable in some people with adenomas.

These findings have prompted the hypothesis that ETBF can contribute early in tumor development. They don't show that stool testing detects risk before a polyp becomes visible at colonoscopy. The pooled cancer-stage analysis found more frequent detection in advanced cancer than early cancer, and that finding changed when influential studies were removed. There is no reliable stage-by-stage pattern that you can apply to an individual result.

Other human studies have found no clear association. A stool study of 156 people found no statistically significant difference in ETBF detection between cancer and control groups. A separate study using colonoscopy tissue found no independent association between bft and precancerous or cancerous lesions. In people with Lynch syndrome, detecting ETBF in an existing tumor wasn't significantly associated with subsequent colorectal cancer. The tissue findings concern a different specimen from this stool test.

Inflammatory Bowel Disease

The evidence in inflammatory bowel disease is inconsistent. One small study using targeted stool PCR found the toxin gene in about one fifth of people with active disease and none of those with inactive disease. A later Crohn's disease study found no significant difference in toxin-gene detection or abundance between active disease and remission.

A study of ulcerative colitis found the gene much more frequently in colon biopsies from affected people than in controls. Those were tissue measurements, and they don't establish the performance of a stool result. Taken together, the studies don't support using this marker to identify the cause of a flare or to monitor inflammatory bowel disease activity.

Inflammatory Diarrhea

ETBF can cause diarrheal illness. Human studies have described abdominal pain, watery diarrhea and intestinal inflammation, often without fever. In one observational study, inflammatory findings persisted after the diarrhea had resolved, including in people who received antibiotics.

Asymptomatic carriage complicates interpretation. A Swedish study detected ETBF in about 27% of adults with diarrhea and 12% of controls. A study of children in Kolkata found the same prevalence in both groups. These studies used targeted testing and, in some cases, cultured isolates or toxin-activity assays. They don't establish that a metagenomic production estimate identifies the cause of your diarrhea.

When Results Can Be Misleading

• Genetic potential and toxin activity differ. Finding the gene doesn't establish how much toxin is being produced. Laboratory studies show that toxin-gene activity changes with bacterial growth conditions.
• A percentage depends on the rest of the community. Relative abundance can change when other organisms increase or decrease, even without a matching change in the absolute number of toxin-carrying bacteria.
• Stool and colon tissue sample different environments. Tissue studies examine bacteria associated with the colon lining. Their detection rates and disease associations cannot be transferred directly to a stool measurement.
• Methods affect detection. DNA extraction, the amount of sequencing, reference databases and analysis methods can change what is detected and how it is quantified. Targeted PCR counts, metagenomic percentages and toxin-activity assays are not interchangeable.
• A negative result has limits. A small population can fall below detection, and the sampled stool may not represent every part of the colon. An undetected result doesn't exclude colon disease.
• Treatment and bowel preparation can change the community. Antibiotics and colonoscopy preparation can alter what is present when you collect the sample. This can represent a real temporary biological change, rather than an analytical error.

Changes Over Time

Microbial carriage can change, and repeated detection can document that a trait remains detectable. But the size or direction of a change in this result has no validated relationship with future cancer, intestinal inflammation or treatment success. There is no established health target for the percentage.